Method and system for a selection optimization process

• US 7,949,553 B1
• Filed: 09/25/2003
• Issued: 05/24/2011
• Est. Priority Date: 09/25/2003
• Status: Active Grant

• Current Assignee: PROS Incorporated (PROS Holdings, Inc.)
• Original Assignee: Pros Revenue Management, L.P.
• Inventors: Wisniewski, Michael, Buraparate, Viroj, Aswal, Navin, Oosten, Maarten
• Primary Examiner(s): Jeanty; Romain
• Assistant Examiner(s): FEENEY, BRETT A
• Application Number:
  US10/670,989
• Time in Patent Office: 2,798 Days
• Field of Search: None
• US Class Current: 705/7.33
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 30/0204 Market segmentation
• CPC Class Codes: G06Q 30/0251 Targeted advertisements
• CPC Class Codes: G06Q 30/0253 During e-commerce, i.e. online transactions
• CPC Class Codes: G06Q 30/0254 based on statistics

Abstract

Systems and methods are described which optimize a business process through adaptive learning. Incoming customers are profiled and based on the customer'"'"'s profile, a corresponding set of actions may be generated. From this set of actions, an algorithm is used to select a discrete action which is then presented to the customer. The customer'"'"'s response to the presented action may in turn be used to forecast future customer responses and update the selection algorithm. These methods and systems allow attainment of a specific objective through progressive optimization of a business process.

Claims

1. A computer-implemented method for the optimization of a process, comprising:

providing a system comprising;

an action selection computer, comprising;

a storage medium for storing an action database containing a set of offers and an action selection module; and

a first processor; and

a customer profiling computer comprising;

a storage medium containing a customer profile database and a customer profiling module comprising at least one customer profiling algorithm; and

a second processor for executing the at least one customer profiling algorithm;

receiving, by the customer profiling module on the customer profiling computer, information about a customer and a customer need;

matching, by the customer profiling module on the customer profiling computer, the information about the customer to a profile stored in the customer profile database based on the customer information;

selecting, by the action selection module on the action selection computer using an action selection algorithm, an offer from a set of offers associated with the profile, wherein each offer in the set offers is unique, wherein the set of offers comprises a plurality of alternative offers associated with the customer need and a business context;

presenting, by the action selection module on the action selection computer, the offer to the customer associated with the profile;

receiving, by the action selection module on the action selection computer, a response to the offer from the customer,

storing the response in the action database in the action selection computer;

repeating the selecting, presenting, receiving and storing steps for the customer and each of a plurality of customers, wherein an order of selection and presentation of offers differs between the plurality of customers;

determining a distribution of customer responses to the set of offers;

analyzing, by the action selection module, the distribution of responses to all offers of the set of offers presented to the customers associated with the profile, wherein the analysis identifies an order of selection and presentation of offers of the set of offers that best fits the customer need and the business context for a given profile; and

updating, by the action selection module, a selection and presentation order of offers of the set of offers in the action database based on the analysis of the distribution of responses, wherein future offers presented to customers associated with the profile are selected by the action selection module from the updated action database based on the analysis,

wherein the order of selection and presentation of offers is an action and different orders of selection and presentation of offers are different actions, wherein actions selected for the customer and each of a plurality of customers is determined by updating a response forecast and updating the action selection algorithm by one of;

updating the response forecast and then updating the action selection algorithm according to the updated response forecast wherein the updated action selection algorithm is applied intermittently by one of;

a per customer basis, a batchwise basis or in an amalgam of the per customer basis and the batchwise basis;

or

updating the response forecast and updating the action selection algorithm by generating a n-tuple vector of proportions, where n defines the number of actions, wherein the n-tuple vector of proportions defines how a stream of profiled customers are to be allocated across the set of actions, randomly allocating customers according to the n-tuple vector of proportions and then iteratively updating the action selection algorithm and the response forecast according to the responses to the set of actions by the stream of profiled customers.

2. The method of claim 1, wherein matching the information about the customer to a profile further comprises collecting historical information.

3. The method of claim 2, wherein receiving information about a customer further comprises augmenting the information with information received from external sources.

4. The method of claim 1, wherein the set of offers is specific to the profile, wherein each offer in the set of offers is selected based on a response forecast, and wherein a response to the offer presented to the customer is used to update the set of offers contained in the response forecast.

5. The method of claim 4, wherein the action database contains a history of responses for each offer, wherein the action selection module analyzes the history of responses associated with the profile in selecting an offer from the set of offers.

10. The method of claim 1, wherein the business context is to maximize profit.

11. The method of claim 1, wherein the business context is to maximize revenue.

12. The method of claim 1, wherein the offer presented to a customer is selected from the set of offers based on the price, wherein the price is below the customer'"'"'s willingness to pay.

6. A system for the optimization of a process, comprising:

a customer profiling computer comprising;

a first processor;

a non-transitory machine readable media storing a customer profile database containing instructions that when executed by the processor configure the customer profiling computer to;

receive information about a customer and a customer need;

match the information about the customer to a profile stored in the customer profile database based on the customer information and one or more of a customer need and a business context; and

an action selection computer comprising;

a second processor; and

a second non-transitory machine readable media storing an action database containing a set of offers and a set of instructions that when executed by the second processor configure the action selection computer to;

select, using an action selection algorithm, an offer from a set of offers associated with the profile, wherein each offer in the set offers is unique, wherein the set of offers comprises a plurality of alternative offers associated with the customer need and a business context;

present the offer to the customer associated with the profile;

receive a response to the offer from the customer,

store the response in the action database in the action selection computer; and

repeat the select, present, receive and store steps for the customer and each of a plurality of customers, wherein an order of selection and presentation of offers differs between the plurality of customers;

determine, by the action selection computer, a distribution of customer responses to the set of offers;

analyze, by the action selection computer, the distribution of responses to all offers of the set of offers presented to the customers associated with the profile, wherein the analysis identifies an order of selection and presentation of offers of the set of offers that best fits the customer need and the business context for a given profile; and

update, by the action selection computer, a selection and presentation order of offers of the set of offers in the action database based on the analysis of the distribution of responses, wherein future offers presented to customers associated with the profile are selected by the action selection module from the updated action database based on the analysis,

wherein the order of selection and presentation of offers is an action and different orders of selection and presentation of offers are different actions, wherein actions selected for the customer and each of a plurality of customers is determined by updating a response forecast and updating the action selection algorithm by one of;

updating the response forecast and then updating the action selection algorithm according to the updated response forecast wherein the updated action selection algorithm is applied intermittently by one of;

a per customer basis, a batchwise basis or in an amalgam of the per customer basis and the batchwise basis;

or

updating the response forecast and updating the action selection algorithm by generating a n-tuple vector of proportions, where n defines the number of actions, wherein the n-tuple vector of proportions defines how a stream of profiled customers are to be allocated across the set of actions, randomly allocating customers according to the n-tuple vector of proportions and then iteratively updating the action selection algorithm and the response forecast according to the responses to the set of actions by the stream of profiled customers.

7. The system of claim 6, wherein the set of offers is specific to the profile, wherein each offer in the set of offers is selected based on a response forecast, and wherein a response to the offer presented to the customer is used to update the set of offers contained in the response forecast.

8. A system for the optimization of a process, comprising:

a customer profiling computer comprising;

a first processor;

a non-transitory machine readable media storing a customer profile database and a customer profiling module that include a set of instructions that when executed by the processor configure the customer profiling computer to perform;

receiving, by the customer profiling module, information about a customer and a customer need;

matching the information about the customer to a profile stored in the customer profile database based on the customer information and one or more of a customer need and a business context; and

an action selection computer comprising;

a second processor; and

a second non-transitory machine readable media storing an action database containing a set of offers; and

an action selection module that include a set of instructions that when executed by the second processor configure the action selection computer to perform;

identifying a set of offers to be presented to a plurality of customers based on the profile associated with the plurality of customers, wherein each offer in the set of offers is unique, wherein the set of offers comprises a plurality of alternative offers associated with the customer need and a business context;

selecting, using an action selection algorithm, an offer from a set of offers;

presenting, by the action selection computer, the offer to the customer associated with the profile;

receiving a response to the offer from the customer,

storing the response in the action database in the action selection computer; and

repeating the selecting, presenting, receiving and storing steps for the customer and each of a plurality of customers, wherein an order of selection and presentation of offers differs between the plurality of customers;

determining a distribution of customer responses to the set of offers;

on the action selection computer, analyzing the distribution of responses to all offers of the set of offers presented to the customers associated with the profile, wherein the analysis identifies an order of selection and presentation of offers of the set of offers that best fits the customer need and the business context for a given profile; and

updating, by the action selection computer, a selection and presentation order of offers of the set of offers in the action database based on the analysis of the distribution of responses, wherein future offers presented to customers associated with the profile are selected by the action selection module from the updated action database based on the analysis,

wherein the order of selection and presentation of offers is an action and different orders of selection and presentation of offers are different actions, wherein actions selected for the customer and each of a plurality of customers is determined by updating a response forecast and updating the action selection algorithm by one of;

updating the response forecast and then updating the action selection algorithm according to the updated response forecast wherein the updated action selection algorithm is applied intermittently by one of;

a per customer basis, a batchwise basis or in an amalgam of the per customer basis and the batchwise basis;

or

updating the response forecast and updating the action selection algorithm by generating a n-tuple vector of proportions, where n defines the number of actions, wherein the n-tuple vector of proportions defines how a stream of profiled customers are to be allocated across the set of actions, randomly allocating customers according to the n-tuple vector of proportions and then iteratively updating the action selection algorithm and the response forecast according to the responses to the set of actions by the stream of profiled customers.

9. The set of instructions of claim 8, wherein the set of offers is specific to the profile, wherein each offer in the set of offers is selected based on a response forecast, and wherein a response to the offer presented to the customer is used to update the set of offers contained in the response forecast.

13. The method of claim 8, wherein the business context is to maximize profit.

14. The method of claim 8, wherein the business context is to maximize revenue.

15. The method of claim 8, wherein a set of offers comprises a set of amenities, wherein each offer contains one or more amenities, wherein the distribution of responses is analyzed to determine which one or more amenities maximizes profit.

16. The method of claim 8, wherein a set of offers comprises a set of amenities, wherein each offer contains one or more amenities, wherein the distribution of responses is analyzed to determine which one or more amenities maximizes revenue.

17. A method comprising:

receiving, by a processor, a set of action for a first customer profile, each action in the set of action including a plurality of offers and a plurality of free amenities corresponding to each of the plurality of offers and an order of presentation of the plurality of offers differs for each action;

receiving, by the processor, a business context;

receiving, by the processor, information about a plurality of customers and needs of the plurality of customers;

determining, by the processor, that the plurality of customers match the first customer profile;

selecting, by the processor, an action from the set of actions for the first customer profile for presentation to each of the plurality of customers wherein the order of selection and presentation of offers is an action and different orders of selection and presentation of offers are different actions, wherein actions selected for the customer and each of a plurality of customers is determined by updating a response forecast and updating the action selection algorithm by updating the response forecast and updating the action selection algorithm by generating a n-tuple vector of proportions, where n defines the number of actions, wherein the n-tuple vector of proportions defines how a stream of profiled customers are to be allocated across the set of actions, randomly allocating customers according to the n-tuple vector of proportions and then iteratively updating the action selection algorithm and the response forecast according to the responses to the set of actions by the stream of profiled customers;

determining, by the processor, a distribution of responses from each of the plurality of customers to the presented action in the set of actions; and

identifying, by the processor, based on the distribution of responses and accounting for the vector of proportions, an action from the set of actions that best fits the customer need and business context for the first customer profile.